Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: ci for a standard deviation


From   Richard Goldstein <[email protected]>
To   [email protected]
Subject   Re: st: ci for a standard deviation
Date   Wed, 26 Feb 2014 08:35:51 -0500

Steve,

thanks for this

I note, however, that (1) the bootstrap CI is narrower than the normal
CI; (2) since I am using this as a "sensitivity" analysis for a sample
size estimation, I prefer the broader CI :-)

Best,
Rich

On 2/25/14, 7:11 PM, Steve Samuels wrote:
> Alan's advice is valid only for normally distributed data. For a general approach, use
> the bootstrap or jackknife:
> 
> *************CODE BEGINS*************
> sysuse auto, clear
> set seed 49891
> bootstrap r(sd) : sum mpg
> estat bootstrap, all
> **************CODE ENDS**************
> 
> Steve
> [email protected]
> 
> 
> On Feb 25, 2014, at 10:08 AM, Richard Goldstein <[email protected]> wrote:
> 
> Alan,
> 
> thank you
> 
> Best,
> Rich
> 
> On 2/25/14, 9:57 AM, Alan Neustadtl wrote:
>> Rich,
>>
>>
>> How about:
>> Assuming sd=15 and n=1,000
>>
>> di 15*sqrt((1000-1)/invchi2((1000-1), .05/2))
>> di 15*sqrt((1000-1)/invchi2((1000-1), 1-(.05/2)))
>>
>> Best,
>> Alan
>>
>> On Tue, Feb 25, 2014 at 9:49 AM, Richard Goldstein
>> <[email protected]> wrote:
>>> Hi,
>>>
>>> I want to calculate a CI for a standard deviation; anyone have Stata
>>> code for this?
>>>
>>> And, no, it is not the CI given by sdtest (that is a CI for the mean for
>>> some reason)
>>>
>>> Rich
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index